Selective nuclease digestion of messenger RNAs inside living cells via the short interfering RNA (siRNA)-triggered RNA interference (RNAi) pathway has become a mainstay in molecular biology to study gene function and holds promise for the development of a powerful and broadly applicable new class of therapeutics. However, specific issues exist that limit the application of this promising technology. Key among these are off-target effects that arise from the ability of siRNAs to interact with proteins involved in the innate immune response and the fact that mRNAs with imperfect matches to the siRNA guide strand are also targeted. In this competitive renewal of an R01 project, we will advance our understanding of these off-target effects and how to control them through the application of synthetic nucleic acid chemistry coupled with techniques from biochemistry and molecular and cellular biology. A challenge in the field of siRNA research is to develop guide strand modifications that will lead to more specific targeting (i.e. reduced tolerance of mismatches in the guide-target duplex). Here we will investigate three different approaches to increase the specificity of Ago-guide + target strand binding. First, we will identify modifications to the guid stand 3' end that modulate affinity for the Ago2 PAZ domain. Disengaging the 3' end from this domain is proposed to be an important step in forming the silencing-competent complex. We propose here that enhancing the interaction between the guide strand 3' end and the PAZ domain will require stable pairing outside the seed for 3' disengagement, rendering the complex less tolerant of mismatches. Second, we will develop new nucleobase analogs that simultaneously recognize two adjacent nucleotides in the target. We propose that with such analogs, a mismatch at either location in the target would destabilize the complex to a greater degree than if standard Watson-Crick complementary bases were used because the destabilizing effect of the mismatch will be translated through the linker. Finally, we will take advantage of the fact that the guide-target duplex forms within a complex with the Ago2 protein. We propose here that nucleobase analogs with appended groups could provide additional destabilization to the mismatched pair and/or stabilization to the matched pair via interactions between the appendage and the protein. Stimulation of the innate immune response by binding Toll-like receptors is another important siRNA off target effect to be addressed in this project. Sequence motifs found in the component strands of certain siRNAs are known to activate TLRs 7 and 8 and induce an immune response in human cells. During the last funding period, we showed that modifications to the nucleobases in immunostimulatory siRNAs reduce cytokine production in human PBMCs. While many different RNA sequences are now known to be immunostimulatory and different chemical modifications shown to block the effect, no concise model has been presented for the ligand/receptor interaction that can explain why certain sequences are favored over others or that can explain the effects of chemical modifications. We will carry out experiments to define further the immune receptor ligand in immunostimulatory siRNAs, to test novel hypotheses for immune receptor binding and to provide new nucleoside analogs that either block TLR activation or increase potency of the RNAi effect. !